Mr. Justice CLARK:
Bargain Car Wash, Inc. appeals from an order of dismissal with prejudice of its treble damage action against American Oil Company and the entry of a judgment for $3466.00 on a cross action against Bargain for conversion of personal property and the balance due on open account. The ease was tried before the court without a jury and these findings were entered: (1) The gasoline sales involved were not in commerce under the provision of the Clayton Act; (2) American’s Trading Area Competitive Allowance [TACA] given its retail gasoline dealers doing business in the same TACA zone to aid them in meeting retail price cuts of competing brands in their zone and other discounts were not violative of Section 2(a) 1 of the Act; (3) the price changes so made in a zone by American were, in any event, made in response to changing conditions affecting the market for gasoline in the zones concerned and caused no injury to Bargain; (4) in any event the allowance of TACA comes within the proviso of § 2(b)2 of the Act; and (5) that Bar[1166]*1166gain’s loss was occasioned by its sales below its costs, its inefficient operation, the marginal character of its location and the ineptness of its management rather than by any action of American.
Our study of the record leaves us with “a firm and definite conviction that a mistake has been committed,” Prince v. Packer Mfg. Co., 419 F.2d 34, 38 (7 Cir. 1969), not only on the commerce ruling but also with reference to other basic findings which leads us to order the dismissal set aside; the judgment on the cross action affirmed but a partial new trial ordered on Bargain’s claims. We find violations present under Section 2(a), remand for a determination of (1) American’s Section 2(b) defense and, if the latter is found not available, (2) then Bargain’s damages.
1. Interstate Commerce:
The gasoline sales involved in this suit are well within the jurisdictional requirements of the Act. The gasoline was refined from crude oil originating in states other than Illinois at refineries located at Sugar Creek, Missouri, and Whiting, Indiana. It was transported by American via its private pipelines to its O’Hare Terminal in Des Plaines, Illinois, its distribution facility for North Side Chicago. The tanks at this terminal have a capacity of some 8.-75 million gallons and are replenished every five to seven days depending upon the gasoline demand of the dealers and customers. The tanks are seldom filled, the gasoline being pumped daily into waiting trucks for delivery to American dealers for re-sale to motorists. Delivery to dealers vary from every day to three days each week, depending on the dealer. Bargain secures all of its gasoline from American and uses American’s Meter Marketing Plan by which the title to the gasoline remains in American until it goes through the retail pump and into the motorist’s car. It is a simultaneous sale from American to the dealer and from the dealer to the customer. This plan shortens the credit period on the payment for the gasoline by the dealer but, of course, has no effect on its interstate character. This Circuit itself has passed upon a similar factual situation as far back as 1949. Standard Oil Company v. FTC, 173 F.2d 210, 213 (7 Cir.), a ruling affirmed by the Supreme Court. See 340 U.S. 231, 237, 71 S.Ct. 240, 95 L.Ed. 239 (1951). The character of the transaction is universally recognized and was, in fact, stipulated by American itself in Sano Petroleum Corporation v. American Oil Co., 187 F.Supp. 345 (E.D.N.Y.1960).
2. The Parties and their Business:
Bargain is an Illinois corporation, Mrs. Robert Deutsch, the wife of its President being its sole stockholder and Secretary. It was the sub-lessee of a gasoline station-car wash facility at 1750 West Foster Avenue on the North side of the City of Chicago a location that was under a long-term lease to American. The sub-lease was effective October 10, 1967, and Bargain began a unitized retail gasoline-car wash business about November 1st. The sub-lease was cancelled by Bargain on December 26, 1968, because of heavy losses. Bargain maintained a seven-day week and a 14-16 hour day. Deutsch worked 6 days each week. The gasoline was purchased from American at tank wagon prices that were normally 17.5$ per gallon for regular and 21.0$ for premium. In addition, American collected 10$ per gallon for. federal and state motor fuel taxes. Suggested Retail Prices [SRP] were posted by American but were entirely optional.
[1167]*1167We need not detail the hazards incident to the retailing of gasoline, oil and their accessories. Suffice it to say that interbrand price competition is intense but intrabrand reaches even a whiter heat for many reasons, including the habits of motorists to usually stick to one brand and the ease of credit card purchases. While prizes and other gimmicks are used as lures by the gasoline suppliers, the dealer also indulges in them in his quest for gallonage. However, the two most effective methods for increasing sales volume is (1) the operation of a unitized gasoline-car wash station and (2) the cutting of the retail price of the gasoline. The former affords a painless way for a customer to get a free car wash instead of a price cut. The thousands of gasoline stations, the high density of their location and the magnitude of the continued competitive efforts of their suppliers keep the industry on uneasy street most of the time and spark many price wars.
3. American’s System of Competition: In 1955-1956 American met the price cutters head-on by inaugurating its Trading Area Competitive Allowance [TACA], American’s TACA system divided the City of Chicago into sales zones. The record does not reveal the number of them but if they are numbered consecutively, they must be in the hundreds because zones with as high a number as 335 are involved here. The findings and the record indicate TACA to be quite a sophisticated system. It works along the following lines. When a lower retail pump price on another brand of gasoline is reported to American usually by its own dealer] it has its salesman in the affected zone inspect the posted pump prices at American and other brand stations in the zone and record them on a small zone map. This is reported to the local district office of American where it is recorded on a similar map. When this data indicates reduced prices at other gasoline brand stations, the pricing clerk telephones his counterpart in the local offices of the oil companies supplying the competitive brand stations involved to determine what their respective tank wagon prices are to their dealers in the relevant zone. This data is recorded on the zone map. The District Office then telephones this information plus the volume losses. of American dealers in the relevant zone, where available, to the American General Office Pricing Department. That office decides whether to reduce the price to American dealers in the relevant zone. The reduction is made in the form of a TACA off the pre-existing tank wagon price and is made only to American dealers in the relevant zone. The TACA system also includes the use of SRP to American dealers which is at a uniform level throughout the City of Chicago. When TACA reduces American’s tank wagon price .6 cent the SRP is reduced 1 cent and for each additional .7 cent reduction in the tank wagon price, the SRP is reduced 1 cent. Hence on the first 1 cent reduction the dealer absorbs .4 cent while for each additional 1 cent, he absorbs only .3 cent. The dealer has no say in American pricing policy and takes no part in the granting or withholding of TACA. The record does not show where the decision to grant TACA is actually made, but it is somewhere in the General Offices of American.
In Finding No. 8 the trial court states that TACA zones were designed by American’s marketing personnel, including its marketing representatives concerned with the geographical areas, its district sales managers, district managers, regional and general office pricing staff and its law department. Each zone is said to cover an area from which American’s dealers, as well as its competitive brand ones, primarily draw their customers. There is no finding as to the creation and development of TACA; nor with regard to the factors, standards or guidelines that were used in determining the original boundaries of the zones; nor was any exhibit or other writing or literature produced thereon. In fact, a high American official testified that the TACA system was devised as a means of assisting Ameri[1168]*1168can dealers to meet any price cuts of a competing brand in a TACA zone. However, two other American officials testified that the purpose of TACA zones was to confine price disturbances to a small area, limit the number of American dealers involved and maintain a higher profit margin on a company-wide basis. Only one witness was produced who had ever participated in the realignment of TACA zones and that was in 1968, soon after this suit was filed. He was not permitted to testify as to the criteria, etc., used. American concedes that when the zone lines were originally drawn that the dealers were not consulted nor their attitude toward the plan ascertained or their expertise sought. It was also undisputed that the zone lines were never changed until some time after this suit was filed. It is readily apparent, as Finding 25 specified, that the zones did not represent actual areas of competition in the sale of gasoline. Indeed, the findings recognize at least twelve zones surrounding Bargain’s zone that were competitive, and American’s witnesses readily admitted as much.3 The apparent secrecy of TACA operation [Deutsch knew nothing of it until after he actually began making sales], the identity of the officials in charge and the criteria used as well as the personalities involved in the making of the final decision kept the details of operation in the dark. Likewise the arbitrary lines of the zones plus the denial of participation by the dealers in their determination casts doubt on their integrity. Indeed, the operation itself is suspect when American’s own brief tells us that “in some instances it [American] did not use a TACA even though it had evidence that competitive brand stations and their suppliers had reduced prices in a zone.”4 That seems to smack of “unsystem.”
Bargain’s experts testified that the TACA system is inherently discriminatory as applied in the area involved in this suit. The arbitrary boundary lines of the zones; the density of American dealers therein and the overlapping competitive trading areas point to this conclusion. While the trial court findings are not clear in this respect, they seem to hold that price discrimination is present but that it is justified because the result does not substantially lessen competition under § 2(a); or the § 2(b) proviso affords American protection; or the damage that Bargain suffered was self-inflicted. In short, a confession and avoidance with the latter being insupportable as we read the record.
4. Bargain’s Area of Competition:
(a) Bargain’s gasoline station-car wash was located in American’s zone 322 in the City of Chicago. It was operated as a unit, generally selling gasoline and a car wash as a package. A free wash was given with each 15-gallon purchase of gasoline originally. In March 1968 this was lowered to 10 gallons but was increased to 25 gallons in July. Coupons were issued which were good at any time. Customers accumulated and used them at any and all times of the year. Gasoline without a wash was sold at a discount below SRP. Bargain’s station was located near the center of a 22-zone area which roughly was some four by five miles in size. Six of the twenty-two zones in the area adjoin Bargain’s No. 322 zone [315, 320, 321, 323, 329, 330], and six additional zones included in Finding 20 are “Zones Somewhat Further Away From Zone 322.” [310, 314, 324, 328, 331, 332], The re[1169]*1169maining ten zones are more remote. The findings do not indicate with clarity the area of competition but witnesses estimated it would be a radius of about IV2 miles of Bargain’s Zone 322. Such an area would include most of the 22 zones. However limiting the competitive area to the 12 enumerated in Finding 20, we note that there are eight American gasoline station-car washes and over 30 American gasoline stations located within the twelve, not including Zone 322 where Bargain was located and which had 2 American stations besides Bargain’s gasoline-car wash. Finding 20 also shows that Bargain had only two TACA’s during 1968, a 1.3 cent TACA from February 6 to March 6 and another 1.30 from April 29 to May 20, 1968, a total of 50 days. However, it also shows that during both these periods, all of the 40 or more American stations in these twelve zones were getting as much or more TACA than Bargain. In addition, four of the adjoining zones got TACA during January before Bargain’s began, and five of the “Somewhat Further Away” zones began their TACA earlier in April.5 This broadcasting of TACA to every zone surrounding Bargain all during the period that it was receiving TACA compounded its problem. It could not undersell any competitor— American or otherwise. The TACA’s were of no value to it whatever in its fight for gallonage. In addition, Bargain suffered the granting of additional TACA’s in adjoining and “somewhat farther away” zones from late June and into August as well as October to the end of the year 1968. Finding 20 shows adjoining Zone 315 with 20 TACA’s in late June through August 6 and from October 11th to November 15th; Zone 323 also adjoining, had a 20 TACA from July 2 to August 6th and October 4th to November 5th; and adjoining Zone 330 had a 20 TACA from June 28th to August 6th and a 1.30 one from September 11th to November 5th. In the zones “Somewhat Further Away,” Zone 310 had a 1.30 TACA from June 13th to August 6th; and 20 TACA’s from August 21st to August 29th and from September 6 to November 5th; Zone 314 had a 20 TACA from June 13 to August 6 and also from September 18 to November 5th; Zone 324 had a 1.30 TACA from June 20th to August 6th and from August 19th to August 29th and a 20 TACA from September 20th to November 5th and November 25th until the end of the year; Zone 328 had 1.30 TACA’s from June 27 to August 6th and September 19th to November 5th; and Zone 331 had a 20 TACA from July to August 27th and a 1.30 TACA from August 27th to 29th and a 20 one from September 25th to the end of the year [except for 22 days in November]; Zone 332 had a 1.30 TACA from June 25 to August 6th and a 20 one from August 19th to August 29th, also from October 17th to November 5th and also from November 27th to the end of the year. Bargain’s Zone 322 had no TACA during any of these periods [June-August and October to December 31] and, of course, Bargain did not. The total number of TACA days given Bargain’s competitors was an average of 135.9 days.
(b) Finding 28 relates that American made lease-backs with certain favored dealers of 1.50 credit per gallon on rent. Finding 29 relates that such rebates were “usual in the industry” and points to Deutsch’s present lease in Brookfield; and Finding 30 identifies two locations where “such rebates” were allowed in 1968 by American; however, in the latter references one of the arrangements was an “advertising allowance,” not a lease-back. In any event the lease-back is a mere subterfuge and a classic example of discrimination. Bargain never received any lease-backs, advertising allowances, rebates or discounts but its predecessor Bradford received a 1-eent rebate per gallon of gas sold. Bargain’s rent was $750 a month, Bradford’s was [1170]*1170$1000 but his one-cent rebate paid him $650 a month in November 1965 and even in 1967 ran $330 each month which reduced his rent below Bargain’s. Furthermore an “advertising allowance” of 1.5 cents per gallon for the first 50,000 gallons and a .5 cent per gallon thereafter were given by American to the gasoline station-car wash at Hermitage and Peterson in 1968. This was a handsome kickback on its 1,200,000 gallons pumped in 1968. Another station at Lincoln and Bryn Mawr not mentioned in the Findings received 1 cent per gallon rebate. American made no showing that justified such rebates and without a showing they are discriminatory.
(c) The record is replete with complaints by Bargain of its failure to receive TACA’s; and concerning American’s discrimination in granting TACA in other zones and in contracting for rebates on advertising and lease-backs. Bargain attributed its low gallonage to these discriminations by American. In the 12 zones surrounding Bargain, the American dealers all received in 1968 an average of 135.9 days of TACA support while Bargain only received 50 days. In addition, the three gasoline station-car wash facilities were getting rebates for advertising and lease-back. For example, the one at Hermitage and Peterson received 158 days of TACA and a 2-cent advertising allowance for 365 days. It sold 1,200,000 gallons in 1968. The record indicates that it was always selling gasoline at 2 cents per gallon less than Bargain. Likewise the dealer at Kimball and Foster in Zone 324 — just a mile and a half west of Bargain — received 213 days of TACA and sold gasoline at 2 cents lower than Bargain.
5. American’s Justification:
The findings say that Bargain’s losses stem from its sales below cost, its inefficient operation, the marginal character of its location and the ineptness of its president, Deutsch, in the gasoline station-car wash business.
(a) Findings 37, 38 and 39 have to do with the average price that Bargain sold its gas. Finding 37 states that Bargain’s average selling price for regular gasoline was 26.8$ and premium as 30.-8$ per gallon. Finding 38 states Bargain lost 1.5$ per gallon on all regular gasoline sold and 1.2$ per gallon on all premium. Finding 39 states Bargain’s selling price “for regular 26.8$ per gallon was at least 3$ below the lowest of the low prices of the other dealers.” We find these findings to be clearly erroneous.6 American’s own Exhibit 36 shows Bargain’s pump prices for regular gasoline ranged from 33.9$ to 41.9$ per gallon and for premium, 37.9$ to 45.9$ per gallon. These prices were frequently greater or equal to American’s own suggested retail price (SRP). However, instead of using these figures which appear to be entirely correct, American had its accountant, Mr. Saathoff, reconstruct what American calls Bargain’s average gasoline sales price. As a base for his calculation, Mr. Saathoff used Plaintiff’s Exhibit 15 which was marked “Income.” As we have stated, Bargain operated its business as a unit. When a customer purchased gasoline at Bargain’s station, he was given a free car wash coupon. Originally a purchase of 15 gallons entitled the purchaser to a free ear wash. Later this was reduced to 10 gallons but was finally increased to 25 gallons. The State of Illinois had a sales tax which was required to be paid on all sales. However a car wash, not being a sale, was not taxable. The State therefore permitted gasoline station-ear wash operators to take into account the free wash operation in reporting to it on its gasoline sales tax. Exhibit 15 reflects the allocation that was made by Bargain’s accountant between gasoline sales and car washes. Mr. Saathoff took Bar[1171]*1171gain’s Exhibit 15 and went to a column headed “Gasoline,” from which he took the figure $96,254.74. He testified that he did not take into account Bargain’s car wash operation which was included in Exhibit 15 in a column marked “Car Wash.” In that column is the figure $56,054.56. Mr. Saathoff did not add the “Car Wash” figure or any part of it in his calculation. The car wash figure was the amount of money allocated by the Bargain accountant to the car wash income for car washes which were given free with the purchases of gasoline. Of course these were properly includable on the total gasoline income. [Most of the car washes produced no income at all since they were included in the price of the gasoline; this “car wash” column in Exhibit 15 was merely a bookkeeping arrangement for state tax purposes.] Unfortunately there is no evidence as to what proportion of the “Car Wash” column includes free coupons and what proportion was paid in cash. Mr. Saathoff, of course, should have obtained this figure. Although we do not have the figure, we do know that most of the income attributable to “Car Wash” in Exhibit 15 came from free coupons given with gasoline purchases and therefore should be included in the gasoline income. Since Mr. Saathoff did not include them, his calculations as to the average price is in error. Mr. Saathoff took the erroneous figure, $96,254.74 — -allowing nothing for the car wash contribution— divided it by Bargain’s 330,000 gallons of sales and came up with an average 28.-80 per gallon. Since Bargain’s gasoline sales were approximately one half premium and one half regular and the differential between premium and regular was 40, Mr. Saathoff came up with an average of 26.80 for regular and 30.80 for premium. As a result this figure is in error since it made no allowance whatever for the ear wash column.
(b) Using the same total sales figures [$96,254.74], Finding 40 states that Bargain’s gross profit “was approximately 5%” of Bargain’s sales. It allowed nothing for the car wash operation. The finding states that 5% “was subnormal: the gross profit for the industry was about 20 to 25% of gross sales.” When the true amount of gross sales [$154,721.-77] is used, Bargain’s gross profit was as high if not higher than that of the industry. The same error was made in Finding No. 41 which holds that Bargain’s payroll was 3% times as large as its gross profits. This is untrue. It also fails to include the car wash operation. Indeed, American’s own accountant [Mr. Voltarel] testified that total sales revenue should be used [$154,721.-77]. If this total had been used Bargain’s payroll was about 30 percent of gross profit which was below industry average. Indeed, Bargain’s payroll was very modest, being only $23,000, of which $15,000 was to Deutsch himself. Finding No. 42 seems to take some umbrage at Bargain allocating $5000 of Deutsch’s total salary [$15,000] to the car wash phase of the business. We find nothing significant in this. Undoubtedly with such a small payroll, Deutsch must have devoted considerable time to the ear wash. He worked six days a week at long hours. The business was a unitary one and the $5000 allocation was again only a bookkeeper’s arrangement. We hold that Findings 40 and 41 are both clearly erroneous.
(c) The trial court’s findings also note the marginal location of the Bargain station as another cause of its demise [Findings 44-51]. Finding 44 relates that Mr. Ricciardi operated the station from 1963-1965. It gives his gallonage at about 50,000 gallons a month which [in 1964 it was 745,095] with his rebate earned a net of $175 to $200 a week. American did not give Bargain a rebate. Finding 45 outlines the management of Mr. Bradford at the station. His volume went up to 65,000 gallons a month by November 1965 [780,000 annually]. The 1966 total was 544,946 gallons. Bradford did not run the business himself, being an absentee owner with his kinfolks actually operating the business. They permitted it to run down in 1966-1967 and when he closed up, it was selling about 33,000 [1172]*1172gallons a month [May 1967]. The station had been closed about four months when Bargain took it over, and as Finding 46 recites, “there was no carryover business from the former operator.” Up to 1966 the station was owned by Beck Oil Company, an American jobber, and both Ricciardi and Bradford were sub-lessees under Beck and each received rebates. In fact, American continued the rebates on the stations it acquired from Beck until the latter’s leases expired. Finding 47 states that industry experience is that a station does better the second year than the first. It records that Bargain sold approximately 334,000 gallons during its year of operation, “slightly less per month” than Bradford. Finding 48 recites that Bargain’s location “was marginal in size hampered by a tight access arrangement to the car wash lane ... a lack of reservoir or back up space for cars entering the ear wash . . . expansion of the location was not practical.” Finding 49 describes the streets, traffic pattern, etc., at Bargain’s location. Finding 50 says motorists approaching the station from the west could not see the station because of an underpass and the street was “barely wide enough to have one lane of traffic. . . . ” Finding 51 advises that American was unable to lease the location to anyone after Bargain cancelled out. All of these findings, drawn by American, are a far cry from what the record shows American told Deutsch in order to induce him to lease the location. In the fall of 1967, Deutsch inquired of Jos. Sullivan, American’s car wash coordinator, as to the availability of a gasoline station-car wash. Donald Butts, American’s sales representative on the north side of Chicago, contacted Deutsch thereafter. The record shows that American represented to Deutsch that the station was in a good location; showed him figures that the location sold between 660,000 and 745,000 gallons of gasoline a year; that the station had potential volume which it estimated at 540,000 for the next 12 months and “was very optimistic that the location would in fact become the success it had been prior to Mr. Bradford’s operation, because it was a very successful car wash” [Sullivan testimony at trial]. Bradford was the absentee operator who placed his relatives in charge of the station, still they sold more gasoline than Bargain.
American made extensive repairs and improvements to the station, put in new lighting and a remodeled office, costing several thousand dollars. It also installed four new dual pumps on Deutsch’s request. At his own expense, Deutsch installed car wash equipment and a cor-relator at Butts’ suggestion. He also erected a 50-foot sign costing $9000 which motorists approaching from the west could see above the railroad viaduct and before entering the underpass. It attracted the motorists approaching the location from the west and improved the visibility of the station from that direction. Bargain remained open seven days a week for some sixteen hours a day, and Deutsch was there for six days of the week. Bargain did not sell tires, batteries and accessories but operated its car wash, as indicated, on a unitized basis giving coupons redeemable at the washer with gasoline purchases.
(d) Findings 53-58 outline Deutsch’s experience in the gasoline station-car wash business. Finding 53 reports his operation in 1964 of a south side gasoline station-car wash [American] selling 840,-000 gallons of gasoline a year. Finding 54 relates his operation in October 1966 of a Belmont Street station [Texaco] where he sold some 60,000 gallons a month until an explosion and fire in March 1967. His loss there was $2900. Finding 56 notes the operation involved in this suit in 1967-1968, while 57 and 58 discuss his present operations. Although American is very critical o,f Deutsch, the record fails to show that he was a failure at the gasoline station-car wash business. Indeed, he has made a big success of the operations he took on after Bargain’s experience with American. A month after cancelling out with American, he opened a gasoline station-car wash in Brookfield where he averaged 38,000 [1173]*1173gallons a month with 4,600 car washes [Pltf.Exh. 67]. In 1970 this operation was selling 43,000 a month with 5,000 car washes [Pltf.Exh. 68], In 1969 he opened up another gasoline station in Niles. [It is known as Tri-State.] It sold 20,000 gallons a month in 1969 [6 months only], but after putting in a car wash it averaged 49,000 gallons a month in 1970 with 4,000 car washes. It is interesting to note that the rent at Brookfield is $600 per month and at TriState only $250. Mr. Mathis, Sales Representative of Union Oil Company [the station at Niles is a Union one], testified that this station was in the upper 5 percent to 10 percent of all Union stations north of Roosevelt Road. As to Deutsch, he testified he is “a good business man and a good dealer.”
In Finding 58 it is reported that Deutsch sold 420,000 gallons in 11 months of 1969 “at its two locations.” This is an error. The record [Exh. 21] clearly shows that the Brookfield station alone sold 420,333 gallons during the 11 months, while Tri-State averaged 20,000 gallons for some 6 months. The finding deducts the 120,000 gallons at Tri-State from Brookfield’s 420,000. This is an error. It should be added, giving Deutsch’s total operation 540,333 gallons, over 200,000 gallons in excess of Bargain’s sales. Finding 58 is clearly erroneous.
6. Other Faulty Findings:
(a) Findings 28-35 have to do with American’s allowances or rebates to dealers. Finding 34 finds that the payment of such allowances or rebates to certain dealers and not to Bargain did not constitute discriminations in price in the sale of gasoline. American cites no reason or authority for this conclusion nor does the court. The rebates cannot be justified on the basis that the owner-dealer pays taxes and insurance, since Bargain paid taxes — other than sales and income — as well as insurance. There is no showing that the recipients were paying more than the going rate for rent without their rebates. Nor is there a showing that the rebates were insubstantial. Indeed, in Bradford’s case American frankly said that his net profit came entirely from the rebate. Nor do we find support for Finding 32 that there was no relation between the rebate and the tank wagon price. Rebates were given on a gallonage basis with greater purchases leading to more rebate. Absent any explanation the rebates were simply discounts on each gallon of gasoline purchased. Since they were not given to Bargain, it follows that they are a discrimination in price. It is “merely a price difference,” as the Court said in Federal Trade Commission v. Anheuser-Busch, Inc., 363 U.S. 536, 549, 80 S.Ct. 1267, 4 L.Ed.2d 1385 (1960) and is therefore price discrimination within the meaning of 2(a). It follows that all of the findings holding to the contrary are clearly erroneous.
(b) Finding 52 recapitulates the causes of Bargain’s failure, i. e., “extremely low gasoline prices . . . low gross profit . . . large payroll expenses . . . lack of carryover business . . . and the undesirable features of 1750 W. Foster.” These have all been canvassed, supra, and found wanting. Since Finding 52 is based on the previous ones, it is likewise clearly erroneous.
(c) Findings 62 and 63 have to do with American’s claim that its TACA allowances, rebates, etc., were made because of declining retail gasoline prices constituting changing conditions in the gasoline market coming within the exemption of Section 2(a) of the Act. Again, neither American’s brief here, nor its findings, which were adopted by the District Court, cites a single case in support of its position. American’s TACA was given to meet a specific cut in price; we see no changed condition in the market from a single price cut in one small American zone in Chicago. The contention falls of its own weight. We might add that this Circuit, as well as others, has held that retail price fluctuations do not constitute “changing conditions” within the meaning of the Sec[1174]*1174tion. See American Oil Co. v. FTC, 325 F.2d 101 (7th Cir. 1963); Moore v. Mead Service Co., 190 F.2d 540, 541 (10th Cir. 1951); Balian Ice Cream Co. v. Arden Farms Co., 231 F.2d 356, 359 (9th Cir. 1955). These findings are, therefore, clearly erroneous.
7. The Section 2(a) Violation and Its Causal Connection:
Requisite to the establishment of a Section 2(a) violation is proof that the price discriminations shown in this case create a reasonable probability of substantial injury to competition. In addition, as Mr. Justice Black wrote for the Court in Perkins v. Standard Oil Company, 395 U.S. 642, 89 S.Ct. 1871, 23 L.Ed.2d 599 (1969), Bargain “must, of course, be able to show a causal connection between the price discrimination in violation of the Act and the injury suffered.” At p. 648, 89 S.Ct. at p. 1874. Let us first discuss the Section 2(a) violation.
(a) While the bare existence of price differentials does not compel an inference of a substantial lessening of competition, the cases go far in that direction. For example, a “substantial” price differential sufficient to “influence retail prices” has been held quite adequate. FTC v. Morton Salt Co., 334 U.S. 37, 49, 68 S.Ct. 822, 92 L.Ed. 1196 (1948); and a price differential without more but in the context of keen competition and tight profit margins furnishes sufficient basis for a conclusion that the requirements of Section 2(a) are present. E. Edelmann & Company v. FTC, 239 F.2d 152, 154 (7th Cir. 1956). Indeed, this Circuit has gone even further holding that “any substantial, sustained [price] differential between competing resellers is prima facie injurious.” National Dairy Products Corp. v. FTC, 395 F.2d 517, 521 (7th Cir. 1968). And the Federal Trade Commission long ago invalidated a differential of ten percent existing between jobbers reselling industrial equipment. Mueller Co., FTC Docket 7514 (Jan. 12, 1962). The actual differentials here are higher than those in Mueller, since a 2 cent TACA equals an 11.4 percent discount and, as to some competitive stations, adding a 2f gallon advertising or rental allowance onto it would double that figure. Furthermore, the duration of the allowances and rebates would bring Bargain’s case within the “prima facie injurious” rule of National Dairy Products Corp., supra.
To sum up, we have here a price differential that meets all of the tests of the cases, i. e., a price differential that influences prices (Morton Salt Company, supra); or a substantial differential that exists in the context of keen competition and tight profit margins (E. Edelmann Company, supra); or a substantial and sustained price differential between competing resellers. (National Dairy Products Corp., supra.)
(b) As Mr. Justice White said in concurring in Perma Life Mufflers, Inc. v. International Parts Corp., 392 U.S. 134, 143, 88 S.Ct. 1981, 1987, 20 L.Ed.2d 982 (1968) “it would be enough with respect to causation if the defendant ‘materially contributed’ to plaintiff’s injury ... or ‘substantially contributed notwithstanding other factors contributed also.’ . Plaintiff need not show that the illegality was a more substantial cause than any other.” See also Continental Ore Co. v. Union Carbide and Carbon Corp., 370 U.S. 690, 702, 82 S.Ct. 1404, 8 L.Ed.2d 777 (1962); Automatic Radio Mfg. Co. v. Ford Motor Co., 390 F.2d 113, 117 (1st Cir. 1968). Bargain must show a causal connection between the price discrimination and the injury it suffered, but this does not require proof of loss of specific sales. Perkins v. Standard Oil of California, supra, 395 U.S. at 644-645; Locklin v. Day-Glo Color Corp., 429 F.2d 873 (7th Cir. 1970); Richfield Oil Corporation v. Karseal Corp., 271 F.2d 709 (9th Cir. 1959). American seems to say that the anticompetitive probabilities of discriminative discounts must be on a permanent basis. We have found no case so holding but, in any event, American’s advertising and rental discounts would fall into the permanent [1175]*1175category, since they ran every day during the year involved here. We believe that the TACA was granted on such a sustained basis, i. e., 135.9 days out of the year in the 12 zones surrounding Bargain and 213 days in Zone 324 that it too would meet the test of the cases. See Continental Baking Co. v. Utah Pie Co., 386 U.S. 685, 87 S.Ct. 1326, 18 L.Ed. 2d 406 (1967); FTC v. Sun Oil Company, 371 U.S. 505, 519, 83 S.Ct. 358, 9 L.Ed. 2d 466 (1963) and American Oil Company v. FTC, 325 F.2d 101 (7th Cir. 1963). But there is other causal connection. Although Bargain’s location was improved at several thousands of dollars of expense and a $9000 sign was erected to lure the customers, the fact is that the gallonage dropped 50 percent below 1965 sales and 40 percent below 1966 ones. There was no change in the movement nor the volume of traffic; the station suffered no labor problems, was in excellent repair and appearance; the management was experienced developing both before and after 1968 a high volume of sales; the location had been recommended by American as a good one with potential sales of 740,000 gallons annually. The only reason left for Bargain’s failure was the TACA and rebates granted by American. The most significant evidence in this regard came from American’s own executives who testified that lower priced dealers would gain volume at the expense of their higher priced competitors and that TACA was essential to competition. Bargain received only 50 days of TACA while its other American competitors got 135.9 days at the least and some of its American competitors also had a 2 cent gallon-age rental or advertising allowance. American has been unable to explain these inconsistencies in its position. It also overlooks the testimony of Bargain’s experts that place the blame right at American’s doorstep. The fact is that one of Bargain’s competitors was an American gasoline station-car wash that built up a million gallon annual sales record while Bargain’s was down 40 percent below the previous year. This competitor enjoyed not only TACA but a rebate besides, neither of which Bargain received. We conclude that the causal connection between the price discrimination in violation of Section 2(a) and the injury Bargain suffered was proven.
8. The Section 2(b) Defense:
American claims, however, that its TACA support or allowances meet the criteria of the Section 2(b) proviso.7 In Sun Oil Company v. FTC, 371 U.S. 505, 83 S.Ct. 358, 9 L.Ed.2d 466 (1963) the competitor of Sun’s dealer had not received any support or allowance from its supplier to enable- it to sell at reduced prices. The Court held that Section 2(b) did not afford Sun any protection. However, it specifically reserved decision where an allowance from Sun’s competitor supplier was present. The Court had not had occasion to pass upon the question. Unfortunately, it faces us and effective administration requires that we answer it. We are inclined to the view that the Section 2(b) defense is available to American if the lower price granted its individual dealer is made to meet an equally low price made by a competitor of American to an individual competitor of American’s dealer 8 This conclusion is unrelated to the zones and TACA’s which American has created and with which we are involved. In fact, the record does not show what American competitor is involved here, nor what American dealers triggered the granting [1176]*1176of the TACA’s. However, we must pass upon the point in order to reach the zone problem in this suit.
The Federal Trade Commission has adopted a statement which is consistent with this conclusion.9 However, in the same statement the Commission discussed the practice of zone pricing, noting that such pricing is illegal under the Robinson-Patman Act “when there is a reasonable possibility that it will injure competition. When the zone is too narrowly delineated and there is a substantial difference in the prices charged competing dealers, injury may occur, i. e., injury to the ability of the nonfavored dealer to compete with the recipient of the discrimination.” Trade Regulation Reports fl 10373, p. 18242. Although the discussion of zone pricing in that statement is in the context of the Commission’s affirmative case, rather than its effect on the assertion of a 2(b) defense, the discussion highlights problems with respect to American’s reliance on the use of its TACA system.
Unless the price concessions granted to competitors of Bargain can be justified on an individual .basis, we believe that the history, background, economic and competitive impact, etc., of the American TACA system must be explored in more detail. American may not sustain its burden of proving that it was merely meeting the equally low price of a competitor simply by proving that its price cuts conformed to its own TACA system unless the record demonstrates a valid economic justification for that system. Was the system developed and has it been used to meet limited competitive situations, or were the zone boundaries arbitrarily drawn to enable American to initiate discrimination rather than merely to defend its business in good faith? On remand, the District Court should develop the history and background o.f the TACA system; the competitive pattern of the zones and the resulting economic effects that the granting of TACA has on competition between American dealers; the rules under which the system is implemented and their effect in the actual experience of the system including the impact on individual dealers, their reaction thereto from a competitive standpoint, and any resulting changes therein. In short, we must know the competitive nature of the system in operation and its impact on individual competing American dealers. Only then can the courts intelligently pass upon the availability of the Section 2(b) defense in a zone operation.
Since the case may be retried on the Section 2(b) defense, as well as the damages, we feel that we should also indicate that the evidence offered under Rule 43 (c) should be admitted and given such weight as it is entitled. We have also considered the cross action of American and find no error in its disposition and the judgment thereon is affirmed.
The judgment of dismissal is therefore vacated and the cause is remanded with instructions that it be restored to the docket; that judgment be entered for Bargain on its Section 2(a) action; [1177]*1177that American’s claim of a Section 2(b) defense be heard and determined, and, if the latter is not sustained, Bargain’s damages be ascertained and judgment entered; and that the final judgment on American’s cross action be re-entered; and
It is so ordered.